A Cross-Product Analysis of Earphone Reviews Using Contextual Topic Modeling and Association Rule Mining

Within the evolving field of sentiment analysis, the integration of topic modeling and association rule mining presents a promising yet underexplored method. This approach currently lacks an organized framework for maximizing insights that aid in drawing robust conclusions concerning customer sentim...

Full description

Saved in:
Bibliographic Details
Main Authors: Ugbold Maidar, Minyoung Ra, Donghee Yoo
Format: Article
Language:English
Published: MDPI AG 2024-12-01
Series:Journal of Theoretical and Applied Electronic Commerce Research
Subjects:
Online Access:https://www.mdpi.com/0718-1876/19/4/170
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850049950358962176
author Ugbold Maidar
Minyoung Ra
Donghee Yoo
author_facet Ugbold Maidar
Minyoung Ra
Donghee Yoo
author_sort Ugbold Maidar
collection DOAJ
description Within the evolving field of sentiment analysis, the integration of topic modeling and association rule mining presents a promising yet underexplored method. This approach currently lacks an organized framework for maximizing insights that aid in drawing robust conclusions concerning customer sentiments. Therefore, this study addresses the need and rationale for having comprehensive sentiment analysis systems by integrating topic modeling and association rule mining to analyze online customer reviews of earphones sold on Amazon. It employs Bidirectional Encoder Representations from Transformers for Topic Modeling (BERTopic), a technique that generates coherent topics by effectively capturing contextual information, and Frequent Pattern Growth (FPGrowth), an efficient association rule mining algorithm used for discovering patterns and relationships in a dataset without candidate generation. This analysis of reviews on ten earphone products identified key customer concerns as follows: sound quality, noise cancellation, durability, and battery life. The results indicate an overall positive sentiment towards sound quality and battery life, mixed reviews on noise cancellation, and significant dissatisfaction with product durability. Using integrated topic modeling and association rule mining offers deeper insights into customer preferences and highlights specific areas for product improvement and guiding targeted marketing strategies. Moreover, we focused on algorithm selection to improve the model’s performance and efficiency, ensuring effective compatibility with our sentiment analysis framework. This study demonstrates how combining advanced data mining techniques and structuring insights from written customer feedback enhances the depth and clarity of sentiment analysis, furthering its applicability in e-commerce research.
format Article
id doaj-art-0305cebe7bc64391a76f9e2565ca53d5
institution DOAJ
issn 0718-1876
language English
publishDate 2024-12-01
publisher MDPI AG
record_format Article
series Journal of Theoretical and Applied Electronic Commerce Research
spelling doaj-art-0305cebe7bc64391a76f9e2565ca53d52025-08-20T02:53:37ZengMDPI AGJournal of Theoretical and Applied Electronic Commerce Research0718-18762024-12-011943498351910.3390/jtaer19040170A Cross-Product Analysis of Earphone Reviews Using Contextual Topic Modeling and Association Rule MiningUgbold Maidar0Minyoung Ra1Donghee Yoo2Department of Computer Science, Mongolia International University, Ulaanbaatar 13330, MongoliaDepartment of Computer Science, Mongolia International University, Ulaanbaatar 13330, MongoliaDepartment of Management Information Systems (Business and Economics Research Institute), Gyeongsang National University, Jinju 52828, Republic of KoreaWithin the evolving field of sentiment analysis, the integration of topic modeling and association rule mining presents a promising yet underexplored method. This approach currently lacks an organized framework for maximizing insights that aid in drawing robust conclusions concerning customer sentiments. Therefore, this study addresses the need and rationale for having comprehensive sentiment analysis systems by integrating topic modeling and association rule mining to analyze online customer reviews of earphones sold on Amazon. It employs Bidirectional Encoder Representations from Transformers for Topic Modeling (BERTopic), a technique that generates coherent topics by effectively capturing contextual information, and Frequent Pattern Growth (FPGrowth), an efficient association rule mining algorithm used for discovering patterns and relationships in a dataset without candidate generation. This analysis of reviews on ten earphone products identified key customer concerns as follows: sound quality, noise cancellation, durability, and battery life. The results indicate an overall positive sentiment towards sound quality and battery life, mixed reviews on noise cancellation, and significant dissatisfaction with product durability. Using integrated topic modeling and association rule mining offers deeper insights into customer preferences and highlights specific areas for product improvement and guiding targeted marketing strategies. Moreover, we focused on algorithm selection to improve the model’s performance and efficiency, ensuring effective compatibility with our sentiment analysis framework. This study demonstrates how combining advanced data mining techniques and structuring insights from written customer feedback enhances the depth and clarity of sentiment analysis, furthering its applicability in e-commerce research.https://www.mdpi.com/0718-1876/19/4/170topic modelingassociation rulessentiment analysistext miningcross-product topicsonline customer reviews
spellingShingle Ugbold Maidar
Minyoung Ra
Donghee Yoo
A Cross-Product Analysis of Earphone Reviews Using Contextual Topic Modeling and Association Rule Mining
Journal of Theoretical and Applied Electronic Commerce Research
topic modeling
association rules
sentiment analysis
text mining
cross-product topics
online customer reviews
title A Cross-Product Analysis of Earphone Reviews Using Contextual Topic Modeling and Association Rule Mining
title_full A Cross-Product Analysis of Earphone Reviews Using Contextual Topic Modeling and Association Rule Mining
title_fullStr A Cross-Product Analysis of Earphone Reviews Using Contextual Topic Modeling and Association Rule Mining
title_full_unstemmed A Cross-Product Analysis of Earphone Reviews Using Contextual Topic Modeling and Association Rule Mining
title_short A Cross-Product Analysis of Earphone Reviews Using Contextual Topic Modeling and Association Rule Mining
title_sort cross product analysis of earphone reviews using contextual topic modeling and association rule mining
topic topic modeling
association rules
sentiment analysis
text mining
cross-product topics
online customer reviews
url https://www.mdpi.com/0718-1876/19/4/170
work_keys_str_mv AT ugboldmaidar acrossproductanalysisofearphonereviewsusingcontextualtopicmodelingandassociationrulemining
AT minyoungra acrossproductanalysisofearphonereviewsusingcontextualtopicmodelingandassociationrulemining
AT dongheeyoo acrossproductanalysisofearphonereviewsusingcontextualtopicmodelingandassociationrulemining
AT ugboldmaidar crossproductanalysisofearphonereviewsusingcontextualtopicmodelingandassociationrulemining
AT minyoungra crossproductanalysisofearphonereviewsusingcontextualtopicmodelingandassociationrulemining
AT dongheeyoo crossproductanalysisofearphonereviewsusingcontextualtopicmodelingandassociationrulemining